:: Re: [DNG] powerdns upstream has dro…
Page principale
Supprimer ce message
Répondre à ce message
Auteur: tito
Date:  
À: dng
Sujet: Re: [DNG] powerdns upstream has dropped sysvinit support
On Fri, 13 Oct 2023 08:44:00 -0300
altoid via Dng <dng@???> wrote:

> On 12 Oct 2023 at 17:48, Steve Litt wrote:
>
> > ... you make me think more than any five other people.
> Individually or at the same time?
>
> [init software]
>
> > Do we need just one window ...
> *Not* the same thing.
> Rather surprised you (seem to) think so.
>
> You can do quite well in Linux without a WM or a desktop.
> And it is quite fitting that you could choose your poison if you
> needed one.
>
> ----
> But a Linux init package is *not* in the same slot/category, it is
> absolutely vital for a Linux box to run *anything* written for it.
> ----
>
> The same way that a desktop has always been vital to any post-DOS MS
> OS.
> Because *any* "Windows" box needed it to run with *windows*.
>
> Which is why MS (have to hand it to them) mastered their desktop
> environment.
> It was their key to the dominance of the desktop PC market.
>
> Well not only that.
> Convincing the general public that "computing" was as easy as turning
> the PC on and clicking on an icon was fundamental.
>
> All that while MS's infamous registry was invented (W95) to steadily
> spread and silently take over the OS, controlling *everything* that
> was going on inside the box without any possibility of knowing what,
> when and why.
>
> [ONE, however imperfect ... ]
> > ... judging from the vast adoption of systemd ...
>
> That is exactly what systemd-less Linux is up against.
> The dominance of the Linux ecosystem by MS/IBM/RH via systemd.
>
> > ... argue about which BSD should rule them all.
> Ah ...
> So you *do* understand.
>
> > ... if I wanted a "take it or leave it" OS ...
> That is *not* what I have said.
> Seems I spoke too soon ...
>
> [Republican loyalists ...]
> > Bikeshedding!
> Exactly.
> One of the most interesting quotes I have ever come across is one
> attributed to a Charles Kettering (1876-1958):
>
> ---
> "If you want to kill any idea in the world, get a committee working
> on it"
> ---
>
> It has been proven to death and is still highly relevant today.
>
> Note that "Committee" must be one of the very few words (in English)
> with three double letters. Sort of tells you something, no?.
>
> > Depends on the definition of "great following" ...
> Right ...
> Now put a committee on getting a definition.
>
> [... anti-bikeshedding principle ...]
> Once you have seen it at work, you cannot unsee it.
>
> You learn to avoid it like the plague because it is a tool in your
> opponent's box.
> And curiously enough, most times *you* end up using it against
> yourself.
>
> > ... at what point does planning become bikeshedding?
> When you put your chances of survival in the hands of a committee.
>
> Today, the chances of Linux systemd-free distributions surviving the
> MS/IBM/RH onslaught are slim at best.
>
> I will say it again: there is a lot of moolah behind the systemd
> pustch.
>
> As time goes by (and in IT time is *very* fast) more and more
> programs will stop using init scripts and as a result, more and more
> programs will not run on systemd-less distributions any more.
>
> [ ... "needed variables ]
> It may be a good idea.
> I am not in a position to argue that point so I won't.
>
> My point is that all efforts/manpower have to be directed at solving
> the problem at hand ie: survival.


This is an attempt to solve the survival problem, that could give us the
manpower that we do not have to migrate/adopt/fork all packages
that used to place files into /etc/init.d/. I recall having read on this
list there were about 1000. By creating something that could be
of use to all other init systems out there which will have the exact
same problem at some point in time (upstream and/or packagers dropping
their files, or need to write them for all daemons out there) we could multiply
the eyes looking at it and the hands working on it.
There is no guarantee that it will work, it is just a chance and it
will help also the init system that you support (sysvinit).
I for sure am not able to make this happen by myself as my knowledge
in IT and programming is self-taught and therefore lacking,
you seem to be experienced in the linux/debian world
so your contribution would be appreciated even if it is
only critics about why this will not work.

> It is *not* being done and the price to pay will be the obsolescence
> of any systemd-less distribution with the net result being the
> absolute control of the Linux ecosystem by MS/IBM/RH.


Exactly.

> To wit:
>
> If you are on a fast sinking ship in the middle of a storm and all
> you find on deck is a huge piece of styrofoam, you hold on to that to
> try to stay afloat.
>
> ---> You *don't* go back below deck to see if you can find a suitable
> life jacket. <---
>
> Having clearly expressed my opinion on the matter, with this post I
> will end my participation in this thread.


Helping Steve in his modular approach that is already one
development step further than my idea if you have the skills
wouldn't be a committee. Have you ever heard of 2 people's committees?

> Thanks for your input.


Ciao,
Tito